I often hear that in the indicated connection is better to NOT use ferules. Is it true?
Maybe, I always use them with flexible cables (the ones with multiple strands).
It’s especially handy for maintenance reasons, without them, everytime you open and close the terminals, there will be some strands cut / lost. doesn’t happen with these guys.
1 Like
Most the time, terminals are designed to work without, but doesn’t hurt either.
If it’s explicit desired for the mppts, then I will change that, musst have overread that.
Just quickly checked the wiring unlimited again.
There is nothing mentioned to not use ferrules, just:
Use ferrules for all wiring connections, especially when connecting to a screw connector that
does not have a cage. If a stranded cable is used without a ferrule the connector screw can
pinch on only a few strands, and the turning motion of the screw can even twist and break
the strands.
For me, this is the equivalent of saying “caged connectors are the exception that dont need it, but it doesn’t hurt either”.
“Use for all” is the dominant part here (for me)
Do you have an (officially) available source, where victron says to omit them on mppts?
Don’t want to discuss about who’s right, but If i would omit them, then I only would do it when beeing able to reference a official doc saying to do so 
Else the more generic “use on ALL connections” would be what I consider the standing order.
1 Like
There was a discussion on this on the old forum, Guy made a clear official statement on the matter of MPPT’s and ferrules.
Unlimited guide is somewhat general and it needs some enhancements particularly around solar.
https://community.victronenergy.com/questions/186714/ferrule-or-not-10-awg-7-strand-pv-wire-into-mppt-r.html
It’s a mystery to me why ferrules are deprecated with the MPPT’s, and just for the MPPT terminals. I’ve only used the regular MPPT’s with terminals (and with ferrules), not the RS unit pictured in the referenced thread. If ferrules don’t work on those terminals, it seems to me there’s something wrong with the terminals.
I read the old thread. Guy Stewart said he would post some resources justifying his arguments, but I did not find them. Were they ever posted?
I will say, absent technical evidence to the contrary, that I suspect the “no ferrules” admonition is wrong, especially the bit about “reduced contact area” (show me measurements of contact resistance, with and without ferrules) - with one proviso: ferrules should always be correctly sized for the wire, and crimped with a proper ferrule crimper.
Perhaps this has something to do with the wire used?
I will add one other thing: ferrules are much more commonly used in Europe and Australasia than in the US. As a result, ferrules are much more readily available for standard European wire sizes (specified in mm^2). Unfortunately while some US AWG wire sizes are a pretty good match in cross-section to European sizes (e.g., 2AWG is about 33.6mm^2, so works well with a 35mm^2 ferrule), others are not (6AWG falls right in between 10 and 16 mm^2 and 8AWG in between 6 and 10mm^2). Ideally, for such wire sizes, proper AWG-sized ferrules (which can be sourced in the US) should be employed.
3 Likes
Just an educated guess…
When in the connection, by tightening, two or more “walls” of the connector are closing to each other in a parallel manner (cage type terminal), then there is no need for a ferule.
This is general true for circuit breakers. Also this will maximize the electrical contact surface.
When in the connection, for tightening purposes, the end of a screw is inserting in a rotational manner, then a ferule is required in order to protect the wires from being cut by the screw rotational movement. Exception on the solid single strand wires, where, obviously, there in no need for a ferule.
4 Likes
Circuit breakers are generally designed to be used with coarse stranded wire. But in marine and other applications, fine stranded wire is employed, and this is less suited to this kind of termination, absent a ferrule.
As for “maximizing contact area”, that is a misunderstanding (reflected also in Guy’s post about MPPT terminals). What counts is the contact resistance achieved, and the long-term stability of that contact resistance, which, for a clamped contact, depends on a combination of surface area and much more critically, contact pressure. More area with less pressure (same clamping force) does not necessarily help, given typical current densities through the contact zone. It is however important that the bulk of the strands carry current.
A ferrule (properly sized and crimped) forms all the strands into a gas-tight bundle after crimping. Then when that enclosed bundle is crushed and properly tightened in a terminal (ideally a cage-type terminal), that bundle is compressed into contact across the final width (enlarged by deformation) and makes a high quality, low resistance contact. The ferrule is ordinarily tinned, so is less subject to corrosion., e.g., with salt air, than if untinned wire is used without a ferrule.
Among the disadvantages of using fine-stranded wire in cage terminals without a ferrule (of course, you should not use such wire with a screw terminal) is that if the wire is significantly smaller that the maximum size allowed for the terminal, the strands on each side of the bundle can get displaced to the sides, while the center of the bundle absorbs the bulk of the clamp force, leaving the outer strands poorly connected. Fine stranded wire used without a ferrule can also be more subject to relaxation of contact pressure over time and with vibration, thermal cycling, etc.
There are of course other advantages for ferrules (I’m assuming the type with a plastic collar), like guarding against errant strands, ease of insertion into the terminal (try inserting the maximum allowed fine-stranded size wire into a terminal without a ferrule, especially in awkward situations), preventing over-insertion of wire stripped too short (thus allowing insulation to enter the clamp area), providing some strain relief to the wire, etc.
9 Likes
With the use of ferrules sometimes the torque for the screw is not followed, instead users just reef on it. A higher torque is needed to sqaush the ferrule into shape. The connection (resistance increases) can run hotter as well if not torqued enough. The spread does help with both issues if it is fine strand.
I would use ferrules for the fine strand if it fails the tug test at the correct torque (so a breaker clamp grip issue)
3 Likes
My 250/85 is running very hot, so I removed the ferrules from the Battery-cabling yesterday, thought maybe resistance here is causing excess heat.
(It didn’t have any impact on temperatures of the mppt, still reaching 45° during noon, measured on the central backside)
However, for the PV wiring, I refused to remove ferrules, puting fine stranded 6mm2 into the monster-brackets without ferrules is way beyond my lets-try-mood 
What I noted however is that the ferrules haven’t been pressed together in a parallel fashion, they more look like a trapezoid after beeing removed from the mppt brackets, so the brackets seem to close more inside than on the front.
I could imagine, that this may be a source for issues, if not enough torque is applied and the ferrules just have a very small contact area because it’s shape resistance has not been overcome
(Where a blank cable would not offer any resistance that may stop it from beeing squeezed in shape)
Exagerated Imagination, little contact area:
Tested with an unused one and only very little “torque”, this outlines what i mean:
If that would have been filled with a cable, and torque would have been reached, only about 50% of the ferrule would have had good contact.
1 Like
The Orion XS has terminals like that and Victron says you can use 25mm² without ferrule.
Most or all manufacturers of cage terminals allow to use them with and without ferrule.
I use ferrules in almost all installations.
It looks cleaner IMO.
It looks like the ferrule in your picture was not crimped with a proper ferrule crimper.
It wasn’t crimped at all, just to outline my observation that mppt brackets do not close parallel.
How long is the ferrule? There is a minimum strip length to get to full depth. Obviously the back part wont be inserted, as the plastic shoulder prevents it from being at the clamp.
That was a short one, didn’t have any long 35mm left, only ordered 4, when I was initially setting up the mppts.
But that should not make a difference with regards to the observation / question that the mppt brackets apparently do not close very parallel, would it?
At least the ferrules pulled out of the mppt where also squeezed about 2mm more towards their end than near the plastic shoulder.
After all, it is just a try / guess to understand why ommitting them could be better / recommended
If not enough Torque is used, or the ferrules are very sturdy, they may just be squeezed in shape on their innermost 5-8 mm, leaving a good portion unsqueezed and not even touching the moving parts of the brackets?
Chances this would happen with plain cabling are about 0, as single strains won’t offer any mechanical resistance that may cause torque beeing reached to early.
The fitting is an industry standard one. It is a rising clamp in most victron mppts. So the clamp moves parrallel. It is physically impossible for it to move any other way.
The only reason physically that i can think of for a wedge shape after clamp is insufficient length of ferrules or insufficient length of wiring in the ferrule so allowing the end to wedge down since there is no force to act on.
The first part of the barrel of the ferrule does not reach the clamp so won’t be squashed at all.
The most ideal set up is the ferrule/wire goes all the way to the back of the fitting for full coverage in the clamp.
If the collar can’t nest in the square flared opening then the barrel has to be even longer.
For non-cage terminals I always use ferrules. For cage terminals they seem optional. The disadvantage beside price and work hours of ferrules is the terminal is limited to one size below against strands without ferrules.
Another thing to consider if using ferrules up from 16mmq: The bigger the size, the more the diffrence between normal strand (H07V-K) and fine stranded (H01N2) as fine strands show a diffrent compact factor with less air between than higher diameter strands. If you use 70mmq H01N2 to connect a Multiplus, the 70mmq DIN ferrules always seems much too big. Same is true for DIN tubular cable lugs used in multi stranded high flexible compacted conductors. This is normally solved by diffrent crimping pliers or intermediate sleeves between conductor and lug. Or installer doesnt take care for this at all and later it gets hot.
Those kind of sleeves are available up from 16mmq and have diffrent dimensions to normal sleeves for H07V-K. You need to have a real good cut with sharp scissors, then carefully strip and slide the sleeve over. Probably this diffrence in sizes is why Guy did not recommend to use (ordinary) sleeves to connect cage terminals. The model number of sleeves for compacted strands at Klauke is VHR and its specifications are available here. They are on stock at many installation material distributors.
Yes, when looking at the clamps without anything inserted, they also seem quite parallel closing.
Except the upper part (on this mppt) is a little loose, and hence shaking a few, maybe a half mm.
So, having weaker resistance towards the end will cause it to tilt a bit.
(Didn’t compare the other two mppts, as they are in use 
)
But most likely my observation is totally neglectable as a standard ferrule would never cause enough mechanical resistance to not have it squeezed in shape for a proper, full length, surface contact with the clamps.
Hi All,
Really good high quality discussion here.
I agree with almost all of your points on both sides.
I have promised in the past to come back with more info, that promise still stands. I proof read an update for wiring unlimited that goes into more details and have just asked if it can be published here first.
I’ll wait for that precise language because for better and worse, what I say on the matter is then interpreted as Victron’s advice.
I am not dogmatic about it. Both with and without can work as long as the principles are understood and the installation is performed accordingly.
However incorrectly installed ferrules, inadequately torqued screws, incorrectly stranded wires are still among the top causes of product damage, and are not a manufacturing fault, but still claimed for warranty.
3 Likes